Molecular characterization and differentiation of Campylobacter fetus subspecies

Summary

The bacterium Campylobacter fetus (C. fetus) can cause disease in humans and animals. In humans, C. fetus can cause septicemia and intestinal illness, mainly in immunocompromised individuals. C. fetus can be isolated from different animal species, including bovine, ovine, goats, birds and cause abortion and infertility in cattle and sheep.

The species C. fetus contain three subspecies; C. fetus subsp. fetus (Cff), C. fetus subsp. venerealis (Cfv) and C. fetus subsp. testudinum (Cft). Cft is recently described and can be isolated from humans and reptiles. This thesis focuses on the mammal-associated C. fetus; Cff and Cfv. 
In 1959, Cff and Cfv are described as different subspecies based on their biochemical and clinical characteristics; strains that were only isolated from the genital tract of cattle and caused infertility in cows were designated subspecies venerealis, and strains that were isolated from the intestines and caused sporadic abortions were designated subspecies fetus. The biochemical tests 1% glycine tolerance and H2S production can differentiate between the two subspecies; Cff strains are 1% glycine tolerant and produce H2S in cysteine-rich medium, and Cfv strains are not 1% glycine tolerant and cannot produce H2S. In 1963, C. fetus strains were isolated that were phenotypically different and able to colonize both the genital tract and intestines of cattle. These strains were not 1% glycine tolerant, but were able to produce H2S, and were designated as a biochemical variant of Cfv; C. fetus subsp. venerealis biovar intermedius (Cfvi). 

Cfv is described as the causative agent of Bovine Genital Campylobacteriosis (BGC), a syndrome characterized by infertility and abortion in cattle. Multiple industrialized countries including the Netherlands have a successful BGC control program. These programs require that bovine semen, embryos and living animals are free of Cfv during import and export. However, BGC is still endemic in many countries, mainly in low and middle income countries. Nowadays, the BGC control programs focus only on Cfv and not Cff, while it is described that Cff isolates can also cause abortions in cattle. It is also known that the biochemical tests used to differentiate Cfv and Cfv are difficult to reproduce and not always reliable. 

Aim of the thesis
The aim of this thesis was to study the molecular characteristics and differences of the mammal-associated Cff and Cfv strains. Furthermore, we studied if we could associate the molecular characteristics with the virulence of the C. fetus strains. 

C. fetus subspecies differentiation
A crucial element in this study was the reliable differentiation and identification of the C. fetus strains. As mentioned above, the biochemical tests are not always reliable and molecular methods are preferable. Multiple molecular methods are described to differentiate Cff and Cfv strains. We tested all available molecular methods on a set of more than 140 C. fetus strains. None of the available methods was able to make a reliable differentiation between the C. fetus strains. To improve the current diagnostics, we developed a new PCR assay to differentiate between Cff and Cfv strains. This PCR assay can be used in all diagnostic laboratories in the world, even laboratories in developing countries, since no advanced equipment is needed. 

Molecular characteristics of the C. fetus strains
Only one closed Cff genome was available at the start of this study. During the study, the genomes of more than 35 C. fetus strains were sequenced and we were able to complete and close the genomes of a Cfvi, Cfv and Cff strain. We determined the molecular characteristics of the selected C. fetus strains with these closed genomes as reference. 

The sequence data allowed us to perform a time analysis with the C. fetus strains, and this analysis revealed that Cfv is recently diversificated from a Cff ancestor, possibly during the improvements in cattle breeding in the 19th century, which may have resulted that this niche-specific clone (Cfv) could spread very fast among the world. 

We performed a phylogenetic analysis on the core genes of the genomes and this analysis divided the strains in two clusters, which were inconsistent with the biochemical characteristics (phenotypes) of the strains. We included more Cff strains and analyzed the single nucleotide polymorphisms (SNPs) of the genomes. This phylogenetic SNP analysis divided the C. fetus strains in five clusters; one cluster with all Cfv and Cfvi strains like the core genes analysis, and four clusters with Cff strains. This division of C. fetus genomes was, like the core genes analysis, not consistent with the phenotypes of the strains. 

We showed in in this study that the biochemical characteristics of the strains were not consistent with the molecular characteristics. However, BGC control programs use the biochemical tests to differentiate between Cff and Cfv, while this differentiation cannot be associated with the virulence of the strains. It is more important for the BGC control programs to differentiate between virulent and non-virulent C. fetus strains. At the start of this study it was already described that Cfv strains can contain a type IV secretion system (T4SS). These systems are involved in the secretion of DNA or proteins of virulent bacterial strains. We studied further which C. fetus genomes contain a T4SS and revealed that the C. fetus genomes can contain multiple T4SSs. These systems were not Cfv-specific, but were also found in Cff stains. 

Conclusion
The results of this thesis showed that the current biochemical tests are not consistent with the molecular characteristics of the C. fetus strains. Phylogenetic analysis of the genomes showed that the C. fetus strains can be divided in at least five different clusters, and showed that Cfv and Cfvi are recently diversificated from Cff. This suggests that Cfv and Cfvi are not different subspecies, but more a Cff clone restricted to the genital tract of cattle. Overall, our results revealed no scientific reason to differentiate the C. fetus strains based on the biochemical tests 1% glycine tolerance and H2S production. We have found several virulence factors in the C. fetus genomes, including T4SSs, but we could not associate these factors with the clinical characteristics of the strains. Further research is necessary to unravel the functionality of the potential C. fetus virulence factors.

Full text